An Integrated Optimization+ Learning Approach to Optimal Dynamic Pricing for the Retailer with Multi-type Customers in Smart Grids. (arXiv:1612.05971v1 [cs.SY])

In this paper, we consider a realistic scenario in the context of smart grids where an electricity retailer serves three different types of customers, i.e., customers with home energy management system embedded in their smart meters (C-HEMS), customers with only smart meters (C-SM), and customers without smart meters (C-NONE). The main objective of this paper is to support the retailer to make optimal day-ahead dynamic pricing decisions subject to realistic market constraints in such a mixed customer pool. To this end, we firstly propose an optimal energy management model for each C-HEMS, two appliance-level customer behaviour learning models for each CSM, and an aggregated demand forecasting model for the whole C-NONE. With the above demand models established capturing energy consumption patterns of different type customers, we then adopt genetic algorithms (GAs) based solution framework to obtain the optimal dynamic prices for the retailer. In addition to smart meters and HEMS, the penetration of Photovoltaic (PV) generation and energy storage in the households is further considered in the pricing optimization model. Simulation results indicate that the proposed pricing optimization algorithms are effective.

from cs.AI updates on arXiv.org http://ift.tt/2hOrJot
via IFTTT